Re: [SI-LIST] : ?Dumb Question Regarding Multilayer Boards

Mark Randol (ryvw50@email.sps.mot.com)
Fri, 27 Mar 1998 10:37:57 -0700

> I'll make a WAG of my own that from your affiliation (Semi-
> conductor Products) that you're talking about test and/or
> eval type boards.

Well, mostly true. The thoughts are that ideally one of our demo
board designs could be used, as is, by a customer. Makes sense.
I think we are still learning exactly what that means. Hence,
my questions.

> My knee-jerk reaction is that if your performance and EMC
> envelopes are currently adequate w/2 layer boards, it may
> be testamonial to the product you're showcasing. "Look, its
> even works on TWO layers!" These two elements, combined
> with density, are the usual motivators for multilayer boards.

True, but that's with VERY careful (and lots of it!) bypassing and
decoupling. Since we haven't checked the emissions from the boards,
I'm reasonably sure there are problems. Perhaps they are fixable
with layout tweaks, maybe not. I've found some SI type problems
that we are trying to fix with layout changes (and might work), so
I don't think my guess is without merit.

> You mention possible space savings only in passing. What
> do you seek to buy with your 1.5-to-2x cost increase if
> you're not driven there by density, performance, or EMC?

My thought is that if the parts count can be reduced, and the
overall size of the board is shrunk, the cost difference may turn out
to be a wash, or better! The reduced complexity should <?> increase
the repeatability of the design too.

> OTOH, it does sound like you are having occaisional per-
> formance issues with the two layer solution. I suspect
> that when those grow to have costs associated with them
> (extra spins, too much time in layout, or too much spent
> on hardware-based shielding) you'll have the justification
> for the jump to multilayer.

Yes, those are the questions that I'm trying to get a handle on.
Might be time to turn on the "shunk works" cloaking device and
just give it a try.

Thanks,

-- 
---------------------------------------------------------------
Mark Randol, RF Measurements Engineer   | Motorola SPS, Inc.
(602)413-8052 Voice                     | M/S EL379
(602)413-4150 FAX                       | 2100 E. Elliot Road
ryvw50@email.sps.mot.com                | Tempe, AZ 85284
---------------------------------------------------------------